2019-01-24 01:54:55

by Nicholas Mc Guire

[permalink] [raw]
Subject: [PATCH V4] livepatch: non static warnings fix

Sparse reported warnings about non-static symbols. For the variables
a simple static attribute is fine - for the functions referenced by
livepatch via klp_func the symbol-names must be unmodified in the
symbol table and the patchable code has to be emitted. The resolution
is to attach __used attribute to the shared statically declared functions.

Signed-off-by: Nicholas Mc Guire <[email protected]>
Suggested-by: Joe Lawrence <[email protected]>
Acked-by: Miroslav Benes <[email protected]>
Link: https://lore.kernel.org/lkml/[email protected]/
---

V2: not all static functions shared need to carry the __noclone
attribute only those that need to be resolved at runtime by
livepatch - so drop the unnecessary __noclone attributes as
well as the Note on __noclone as suggested by Joe Lawrence
<[email protected]> - thanks !

V3: fix the wording as proposed by Joe Lawrence
<[email protected]> to address that this is not only
about how to fix sparse warnings but also to ensure
traceable/patchable code still being emitted.

V4: fix up the Link to point to the proper page as suggested
by Joe Lawrence <[email protected]>.

Sparse reported the following findings in 5.0-rc3:

CHECK samples/livepatch/livepatch-shadow-mod.c
samples/livepatch/livepatch-shadow-mod.c:99:1: warning: symbol 'dummy_list' was not declared. Should it be static?
samples/livepatch/livepatch-shadow-mod.c:100:1: warning: symbol 'dummy_list_mutex' was not declared. Should it be static?
samples/livepatch/livepatch-shadow-mod.c:107:23: warning: symbol 'dummy_alloc' was not declared. Should it be static?
samples/livepatch/livepatch-shadow-mod.c:132:15: warning: symbol 'dummy_free' was not declared. Should it be static?
samples/livepatch/livepatch-shadow-mod.c:140:15: warning: symbol 'dummy_check' was not declared. Should it be static?

CHECK samples/livepatch/livepatch-shadow-fix1.c
samples/livepatch/livepatch-shadow-fix1.c:74:14: warning: symbol 'livepatch_fix1_dummy_alloc' was not declared. Should it be static?
samples/livepatch/livepatch-shadow-fix1.c:116:6: warning: symbol 'livepatch_fix1_dummy_free' was not declared. Should it be static?

CHECK samples/livepatch/livepatch-shadow-fix2.c
samples/livepatch/livepatch-shadow-fix2.c:53:6: warning: symbol 'livepatch_fix2_dummy_check' was not declared. Should it be static?
samples/livepatch/livepatch-shadow-fix2.c:81:6: warning: symbol 'livepatch_fix2_dummy_free' was not declared. Should it be static?

Patch was compile tested with: x86_64_defconfig + FTRACE=y
FUNCTION_TRACER=y, SAMPLES=y, LIVEPATCH=y SAMPLE_LIVEPATCH=m
(looks sparse, smatch claan, one coccichek warning left - fix later today)

Patch was runtested with:
insmod samples/livepatch/livepatch-shadow-mod.ko
insmod samples/livepatch/livepatch-shadow-fix1.ko
insmod samples/livepatch/livepatch-shadow-fix2.ko
echo 0 > /sys/kernel/livepatch/livepatch_shadow_fix2/enabled
echo 0 > /sys/kernel/livepatch/livepatch_shadow_fix1/enabled
rmmod livepatch-shadow-fix2
rmmod livepatch-shadow-fix1
rmmod livepatch-shadow-mod
and dmesg output compared to previous run.

Patch is against 5.0-rc3 (localversion-next is next-20190123)

samples/livepatch/livepatch-shadow-fix1.c | 4 ++--
samples/livepatch/livepatch-shadow-fix2.c | 4 ++--
samples/livepatch/livepatch-shadow-mod.c | 11 ++++++-----
3 files changed, 10 insertions(+), 9 deletions(-)

diff --git a/samples/livepatch/livepatch-shadow-fix1.c b/samples/livepatch/livepatch-shadow-fix1.c
index a5a5cac..67a73e5 100644
--- a/samples/livepatch/livepatch-shadow-fix1.c
+++ b/samples/livepatch/livepatch-shadow-fix1.c
@@ -71,7 +71,7 @@ static int shadow_leak_ctor(void *obj, void *shadow_data, void *ctor_data)
return 0;
}

-struct dummy *livepatch_fix1_dummy_alloc(void)
+static struct dummy *livepatch_fix1_dummy_alloc(void)
{
struct dummy *d;
void *leak;
@@ -113,7 +113,7 @@ static void livepatch_fix1_dummy_leak_dtor(void *obj, void *shadow_data)
__func__, d, *shadow_leak);
}

-void livepatch_fix1_dummy_free(struct dummy *d)
+static void livepatch_fix1_dummy_free(struct dummy *d)
{
void **shadow_leak;

diff --git a/samples/livepatch/livepatch-shadow-fix2.c b/samples/livepatch/livepatch-shadow-fix2.c
index 52de947..91c21d5 100644
--- a/samples/livepatch/livepatch-shadow-fix2.c
+++ b/samples/livepatch/livepatch-shadow-fix2.c
@@ -50,7 +50,7 @@ struct dummy {
unsigned long jiffies_expire;
};

-bool livepatch_fix2_dummy_check(struct dummy *d, unsigned long jiffies)
+static bool livepatch_fix2_dummy_check(struct dummy *d, unsigned long jiffies)
{
int *shadow_count;

@@ -78,7 +78,7 @@ static void livepatch_fix2_dummy_leak_dtor(void *obj, void *shadow_data)
__func__, d, *shadow_leak);
}

-void livepatch_fix2_dummy_free(struct dummy *d)
+static void livepatch_fix2_dummy_free(struct dummy *d)
{
void **shadow_leak;
int *shadow_count;
diff --git a/samples/livepatch/livepatch-shadow-mod.c b/samples/livepatch/livepatch-shadow-mod.c
index 4aa8a88..4d79c6dc 100644
--- a/samples/livepatch/livepatch-shadow-mod.c
+++ b/samples/livepatch/livepatch-shadow-mod.c
@@ -96,15 +96,15 @@ MODULE_DESCRIPTION("Buggy module for shadow variable demo");
* Keep a list of all the dummies so we can clean up any residual ones
* on module exit
*/
-LIST_HEAD(dummy_list);
-DEFINE_MUTEX(dummy_list_mutex);
+static LIST_HEAD(dummy_list);
+static DEFINE_MUTEX(dummy_list_mutex);

struct dummy {
struct list_head list;
unsigned long jiffies_expire;
};

-noinline struct dummy *dummy_alloc(void)
+static __used noinline struct dummy *dummy_alloc(void)
{
struct dummy *d;
void *leak;
@@ -129,7 +129,7 @@ noinline struct dummy *dummy_alloc(void)
return d;
}

-noinline void dummy_free(struct dummy *d)
+static __used noinline void dummy_free(struct dummy *d)
{
pr_info("%s: dummy @ %p, expired = %lx\n",
__func__, d, d->jiffies_expire);
@@ -137,7 +137,8 @@ noinline void dummy_free(struct dummy *d)
kfree(d);
}

-noinline bool dummy_check(struct dummy *d, unsigned long jiffies)
+static __used noinline bool dummy_check(struct dummy *d,
+ unsigned long jiffies)
{
return time_after(jiffies, d->jiffies_expire);
}
--
2.1.4



2019-01-24 17:06:56

by Joe Lawrence

[permalink] [raw]
Subject: Re: [PATCH V4] livepatch: non static warnings fix

On 1/23/19 8:48 PM, Nicholas Mc Guire wrote:
> Sparse reported warnings about non-static symbols. For the variables
> a simple static attribute is fine - for the functions referenced by
> livepatch via klp_func the symbol-names must be unmodified in the
> symbol table and the patchable code has to be emitted. The resolution
> is to attach __used attribute to the shared statically declared functions.
>
> Signed-off-by: Nicholas Mc Guire <[email protected]>
> Suggested-by: Joe Lawrence <[email protected]>
> Acked-by: Miroslav Benes <[email protected]>
> Link: https://lore.kernel.org/lkml/[email protected]/

Hi Nicholas, thanks for re-posting this fix, the __used attribute change
was particularly interesting to learn about.

I think Miroslav requested a re-ordering of these tags, perhaps we could
do the shuffle when we apply the patch to the tree?

Link:
Suggested-by:
Signed-off-by:
Acked-by:

With that,

Acked-by: Joe Lawrence <[email protected]>


> ---
>
> V2: not all static functions shared need to carry the __noclone
> attribute only those that need to be resolved at runtime by
> livepatch - so drop the unnecessary __noclone attributes as
> well as the Note on __noclone as suggested by Joe Lawrence
> <[email protected]> - thanks !
>
> V3: fix the wording as proposed by Joe Lawrence
> <[email protected]> to address that this is not only
> about how to fix sparse warnings but also to ensure
> traceable/patchable code still being emitted.
>
> V4: fix up the Link to point to the proper page as suggested
> by Joe Lawrence <[email protected]>.

Credit to Miroslav for these last two change suggestions.

-- Joe

> Sparse reported the following findings in 5.0-rc3:
>
> CHECK samples/livepatch/livepatch-shadow-mod.c
> samples/livepatch/livepatch-shadow-mod.c:99:1: warning: symbol 'dummy_list' was not declared. Should it be static?
> samples/livepatch/livepatch-shadow-mod.c:100:1: warning: symbol 'dummy_list_mutex' was not declared. Should it be static?
> samples/livepatch/livepatch-shadow-mod.c:107:23: warning: symbol 'dummy_alloc' was not declared. Should it be static?
> samples/livepatch/livepatch-shadow-mod.c:132:15: warning: symbol 'dummy_free' was not declared. Should it be static?
> samples/livepatch/livepatch-shadow-mod.c:140:15: warning: symbol 'dummy_check' was not declared. Should it be static?
>
> CHECK samples/livepatch/livepatch-shadow-fix1.c
> samples/livepatch/livepatch-shadow-fix1.c:74:14: warning: symbol 'livepatch_fix1_dummy_alloc' was not declared. Should it be static?
> samples/livepatch/livepatch-shadow-fix1.c:116:6: warning: symbol 'livepatch_fix1_dummy_free' was not declared. Should it be static?
>
> CHECK samples/livepatch/livepatch-shadow-fix2.c
> samples/livepatch/livepatch-shadow-fix2.c:53:6: warning: symbol 'livepatch_fix2_dummy_check' was not declared. Should it be static?
> samples/livepatch/livepatch-shadow-fix2.c:81:6: warning: symbol 'livepatch_fix2_dummy_free' was not declared. Should it be static?
>
> Patch was compile tested with: x86_64_defconfig + FTRACE=y
> FUNCTION_TRACER=y, SAMPLES=y, LIVEPATCH=y SAMPLE_LIVEPATCH=m
> (looks sparse, smatch claan, one coccichek warning left - fix later today)
>
> Patch was runtested with:
> insmod samples/livepatch/livepatch-shadow-mod.ko
> insmod samples/livepatch/livepatch-shadow-fix1.ko
> insmod samples/livepatch/livepatch-shadow-fix2.ko
> echo 0 > /sys/kernel/livepatch/livepatch_shadow_fix2/enabled
> echo 0 > /sys/kernel/livepatch/livepatch_shadow_fix1/enabled
> rmmod livepatch-shadow-fix2
> rmmod livepatch-shadow-fix1
> rmmod livepatch-shadow-mod
> and dmesg output compared to previous run.
>
> Patch is against 5.0-rc3 (localversion-next is next-20190123)
>
> samples/livepatch/livepatch-shadow-fix1.c | 4 ++--
> samples/livepatch/livepatch-shadow-fix2.c | 4 ++--
> samples/livepatch/livepatch-shadow-mod.c | 11 ++++++-----
> 3 files changed, 10 insertions(+), 9 deletions(-)
>
> diff --git a/samples/livepatch/livepatch-shadow-fix1.c b/samples/livepatch/livepatch-shadow-fix1.c
> index a5a5cac..67a73e5 100644
> --- a/samples/livepatch/livepatch-shadow-fix1.c
> +++ b/samples/livepatch/livepatch-shadow-fix1.c
> @@ -71,7 +71,7 @@ static int shadow_leak_ctor(void *obj, void *shadow_data, void *ctor_data)
> return 0;
> }
>
> -struct dummy *livepatch_fix1_dummy_alloc(void)
> +static struct dummy *livepatch_fix1_dummy_alloc(void)
> {
> struct dummy *d;
> void *leak;
> @@ -113,7 +113,7 @@ static void livepatch_fix1_dummy_leak_dtor(void *obj, void *shadow_data)
> __func__, d, *shadow_leak);
> }
>
> -void livepatch_fix1_dummy_free(struct dummy *d)
> +static void livepatch_fix1_dummy_free(struct dummy *d)
> {
> void **shadow_leak;
>
> diff --git a/samples/livepatch/livepatch-shadow-fix2.c b/samples/livepatch/livepatch-shadow-fix2.c
> index 52de947..91c21d5 100644
> --- a/samples/livepatch/livepatch-shadow-fix2.c
> +++ b/samples/livepatch/livepatch-shadow-fix2.c
> @@ -50,7 +50,7 @@ struct dummy {
> unsigned long jiffies_expire;
> };
>
> -bool livepatch_fix2_dummy_check(struct dummy *d, unsigned long jiffies)
> +static bool livepatch_fix2_dummy_check(struct dummy *d, unsigned long jiffies)
> {
> int *shadow_count;
>
> @@ -78,7 +78,7 @@ static void livepatch_fix2_dummy_leak_dtor(void *obj, void *shadow_data)
> __func__, d, *shadow_leak);
> }
>
> -void livepatch_fix2_dummy_free(struct dummy *d)
> +static void livepatch_fix2_dummy_free(struct dummy *d)
> {
> void **shadow_leak;
> int *shadow_count;
> diff --git a/samples/livepatch/livepatch-shadow-mod.c b/samples/livepatch/livepatch-shadow-mod.c
> index 4aa8a88..4d79c6dc 100644
> --- a/samples/livepatch/livepatch-shadow-mod.c
> +++ b/samples/livepatch/livepatch-shadow-mod.c
> @@ -96,15 +96,15 @@ MODULE_DESCRIPTION("Buggy module for shadow variable demo");
> * Keep a list of all the dummies so we can clean up any residual ones
> * on module exit
> */
> -LIST_HEAD(dummy_list);
> -DEFINE_MUTEX(dummy_list_mutex);
> +static LIST_HEAD(dummy_list);
> +static DEFINE_MUTEX(dummy_list_mutex);
>
> struct dummy {
> struct list_head list;
> unsigned long jiffies_expire;
> };
>
> -noinline struct dummy *dummy_alloc(void)
> +static __used noinline struct dummy *dummy_alloc(void)
> {
> struct dummy *d;
> void *leak;
> @@ -129,7 +129,7 @@ noinline struct dummy *dummy_alloc(void)
> return d;
> }
>
> -noinline void dummy_free(struct dummy *d)
> +static __used noinline void dummy_free(struct dummy *d)
> {
> pr_info("%s: dummy @ %p, expired = %lx\n",
> __func__, d, d->jiffies_expire);
> @@ -137,7 +137,8 @@ noinline void dummy_free(struct dummy *d)
> kfree(d);
> }
>
> -noinline bool dummy_check(struct dummy *d, unsigned long jiffies)
> +static __used noinline bool dummy_check(struct dummy *d,
> + unsigned long jiffies)
> {
> return time_after(jiffies, d->jiffies_expire);
> }
>

2019-01-25 15:45:06

by Jiri Kosina

[permalink] [raw]
Subject: Re: [PATCH V4] livepatch: non static warnings fix

On Thu, 24 Jan 2019, Nicholas Mc Guire wrote:

> Sparse reported warnings about non-static symbols. For the variables
> a simple static attribute is fine - for the functions referenced by
> livepatch via klp_func the symbol-names must be unmodified in the
> symbol table and the patchable code has to be emitted. The resolution
> is to attach __used attribute to the shared statically declared functions.
>
> Signed-off-by: Nicholas Mc Guire <[email protected]>
> Suggested-by: Joe Lawrence <[email protected]>
> Acked-by: Miroslav Benes <[email protected]>
> Link: https://lore.kernel.org/lkml/[email protected]/

I've reordered the tags :) and applied. Thanks,

--
Jiri Kosina
SUSE Labs


2019-01-27 10:09:11

by Nicholas Mc Guire

[permalink] [raw]
Subject: Re: [PATCH V4] livepatch: non static warnings fix

On Fri, Jan 25, 2019 at 04:44:18PM +0100, Jiri Kosina wrote:
> On Thu, 24 Jan 2019, Nicholas Mc Guire wrote:
>
> > Sparse reported warnings about non-static symbols. For the variables
> > a simple static attribute is fine - for the functions referenced by
> > livepatch via klp_func the symbol-names must be unmodified in the
> > symbol table and the patchable code has to be emitted. The resolution
> > is to attach __used attribute to the shared statically declared functions.
> >
> > Signed-off-by: Nicholas Mc Guire <[email protected]>
> > Suggested-by: Joe Lawrence <[email protected]>
> > Acked-by: Miroslav Benes <[email protected]>
> > Link: https://lore.kernel.org/lkml/[email protected]/
>
> I've reordered the tags :) and applied. Thanks,
>

thx for your patience - I?ll try getting this streight a bit faster next time

thx!
hofrat